Benzazepinone calcium channel blockers. 5 ... - ACS Publications

Benzazepinone calcium channel blockers. 5. Effects on antihypertensive activity associated with N1 and aromatic substituents. Jagabandhu Das, David M...
1 downloads 0 Views 1MB Size
Download PDF
2610

J. Med. Chem. 1992,35, 2610-2617

Benzazepinone Calcium Channel Blockers. 5. Effects on Antihypertensive Activity Associated with N1 and Aromatic Substituents Jagabandhu Das,* David M. Floyd,* S. David Kimball, Keith J. Duff, Michael W. Lago,John Krapcho, Ronald E. White, Richard E. Ridgewell, Mary T. Obermeier, Suzanne Moreland, Diane McMullen, Diane Normandin, S. Anders Hedberg, and Thomas R. Schaeffer Brbtol-Myers Squibb Pharmaceutical Research Institute, P.O.Box 4000,Princeton, New Jersey 08543-4000. Received November 13,1991 We have shown that the pyrrolidinylmethyl substituent on the lactam nitrogen (Nl) of benzazepinone and benzothiazepinone calcium channel blocking agents is resistant to metabolic deamination and generally increasee the duration and potency of antihypertensive activity in spontaneously hypertensive rata (SHR) relative to (NJV-dimethylamino)ethyl analogs. Additionally, compounds poseeseing a substituent on the fused aromatic ring are more resistant to metabolic deacylation of the C3 hydroxy function, which may explain why aromatic substituenta also frequently increase the potency and/or duration of antihypertensive activity. Our data also indicate the increased antihypertensive activity associated with these structural modifications is independent of any effects of potency in vitro. Overall, we interpret these results to indicate that these structural modifications improve antihypertensive activity as a result of increased metabolic stability and, consequently, oral bioavailability.

Introduction We have previously described structureactivity studies with a series of benzazepinone calcium channel blocking agents (1) (X= CHz) which are structural analogs of the benzothiazepinone diltiazem (le). This work resulted in the discovery of compounds that possessed potent antihypertensive activity when administered orally to spontaneously hypertensive rats (SHR).l As a means of focusing our efforts on the discovery of compounds with an extended duration of action, we investigated the metabolic profile of these compounds. As shown in Scheme I, we found the major metabolic pathways involved oxidative dealkylation of the (N,N-dimethylamino)ethyl substituent on the lactam nitrogen (N1 using benzazepinone numbering) to give 2 and 3 and deacylation of the C3 acetoxy functionality giving 4, in accord with data published for diltiazem We found that secondary amines such as 2, resulting from oxidative demethylation of the N,Ndimethylaminogroup, maintained potent calcium channel

Scheme I. General Metabolic Pathways /%OMe

/ /

E.; Ridgewell, R. E.;Moreland, S.; Brittain, R.J.; Normandin, D. E.; Hedberg, S. k;Cucinotta, G. G. Benzazepinone Calcium Channel Blockers. 2. Structure-Activity and Drug Metabolism Studies Leading to Potent Antihypertensive Agents. Comparison with Benzothiazepinones. J. Med. Chem. 1992, 35, 756-772. (2) a. Sugawara, Y.; Ohashi,M.; Nakamura, S.; Usuki, S.; Suzuki, T.; Ito, Y.; Kume, T.; Harigaya, S.; Nakao, A,; Gaino, M.; Inoue, H. Metabolism of Diltiazem. I. Structures of New Acidic and Basic Metabolites in Rat, Dog and Man. J. Pharmacobio-Dyn. 1988,lI,211-223. b. Sugawara, Y.; Nakamura, S.; Usuki, S.; Ito, Y.; Suzuki,T.; Ohashi, M.; Harigaya, S. Metabolism of Diltiazem. 11. Metabolic Profile in Rat, Dog and Man. J. Phurmacobio-Dyn. 1988,lI,224-233. c. Sugihara, J.; Sugaware, Y.;Ando, H.;Harigaya, S.;Etoh, A.; Kohno, K. Studies on the Metabolism of Diltiazem in Man. J. Pharmacobio-Dyn. 1984,7,U-32. d. Montamat, S.C.; Abemethy, D. R.;Mitchell, J. R. High-performance Liquid Chromatographic Determination of Diltiazem and Its Major Metabolites, Nmonodemethyldilthm and Desacetyldiltiazem,in Plasma J. Chromatography 1987,415,203-207.e. LeBoeuf, E.; GrechBelanger, 0. Deacetylation of Diltiazem by Rat Liver. Drug Metab. Dispos. 1987,15,122-126.f. Montamat, S.C.; Abrnethy, D. R.N-monodesmethyldiltiazemis the Predominant Metabolite of Diltiazem in the Plasma of Young and Elderly Hypertensives. Br. J. Clin. Phurmacol. 1987,24,186-189.g. Nakamura, 5.;Suzuki,T.; Sugawara, Y.; Usuki, S.; Ito, Y.; Kume, T.; Yoahikawa, M.; Endo, H.; Ohashi,M.; Harigaya, S. Metabolic Fate of Diltiazem. Arzneim. Forsch. Drug Res. 1987,37,1244-1252.

OM@

m CHO CHIOH C 4 H

NM.,

1 ~~

(1) Floyd, D. M.; Kimball, S. D.; Krapcho, J.; Das, J.; Turk, C. F.; Moquin, R.V.; Lago, M. W.; Duff, K. J.; Lee, V. G.; White, R.

2

\

3

NM.,

4

blocking activity ani. appeared to be resistant to further metabolic degradation. We also observed the alternate oxidative dealkylation process resulting in the mixture of compounds 3. The desamino metabolites 3 were devoid of calcium channel blocking activity due to the removal of the basic amino functionality from the N1 substituent3 These studies led the way to compounds such as 6 (X= CHJ as long-acting and potent antihypertensive agents in the SHR. The key feature of 5 is the pyrrolidinyhethyl N1 substituent, which proved to be resistant to oxidative (3) Kimball,S. D.; Floyd, D. M.; Dee, J.; Hunt, J. T.; Krapcho, J.;

bvnyak, G.; Duff, K. J.; Lee, V. G.; Moquin, R. V.;Turk, C. F.; Hedberg, S. A,; Moreland, S.; Brittain, R J.; McMullen, D. M.; Normandin, D. E.; Cucinotta, G. G.Benzazepinone Calcium Channel Blockers. 4. Structure-Activity Overview and Intracellular Binding Site. J. Med. Chem. 1992,35,780-793.

0022-2623/92/1835-2610$03.00/00 1992 American Chemical Society

Journal of Medicinal Chemistry, 1992, Vol. 36, No.14 2611

Benzazepinone Calcium Channel Blockers

Table I. Physical Properties of Benzazepinones and Benzothiazapinones 5, 10, and lf [alD? % yield compd deg X R1 & from 6 mp, OC recryat solvent H 15* 80-84 CHZ C1 foam Sa +148 H 186 74-78 foam CH2 C1 loa +130 219-21 acetone/pentane 57' 5b +80.5 CH2 CF3 H 10b +130 CH2 CF3 H 75' 165-67 EhO 3od 287-88 AcCN CH2 H OMe 5c +lo3 2od 179-81 EhO/EtOAc CH2 H OMe 1Oc +lo9 H OMe 58' 145-46 AcCN 5d +50.3 S H OMe 1Od +62.9 S 65' 121-24 EhO H H 68' 50 +66.1 S 132-35 EhO >240 EhO/CH2C12 H H 32' 1Oe +91.6 S If +117 CH2 H C1 NA' 63-65 EgO o c = 1, MeOH. *Prepared according to Method I. 'Prepared according to Method 11. mental Section). 'See Experimental Section.

-

Scheme 11." Synthesis of Compounds 5 and 10 a

QCHzOH I

b or c

@CHIOH I

H

formus. anal. CnH&lN203*0.42HzO C,H,N,Cl CuH&lN204.0.29HzO C,H,N,Cl CzaHnFSN2O4~HC1-O.O5HzO C,H,N,Cl,F CaH&'~N20~.HC1*0.73H20 C,H,N,Cl,F C ~ ~ 2 0 ~ * H C 1 . 0 . 2 S H 2 0 C,H,N,Cl CuH&z04.HC1.0.48HzO C,H,N,Cl CuH&O&HCl C,H,N,Cl,S C~H~N204S.HC1.2H20 C,H,N,Cl,S C~H&20&HC1.0.73HzO C,H,N,Cl,S Cz~H~N20&1.1HC1*0.48H20 C,H,N,Cl,S C~H&204.HC1*1.5H20 C,H,N dPrepared according to Method 111 (see Experi-

BOC

o C H 2 R I

BOC

7a; R = OTs 7b; R = Br

c;.'OMe

RZ

OH

N H

OMe

d 7a or 7b

OH

8

For 5,6,8,9,10

X

e

-

O

6

a b c d

OMe

9

f

f

Bi Bz

CH, C1 CH2 CF, CH, H S H S H

H H OMe OMe H

5

10

(a) (BOC),O, CH2C12;(b)p-TsC1, Py; (c) CBr,, PhSP, EbO;(d) NaH, DMF, 80 OC; or Cs2C03,DMF, 50 O C ; (e) Ac20, D W , CH2C12;(f) (i) TFA, CH2C12;(ii) EhO, HC1.

metabolism and resulted in compounds with very high oral bioavailability in the rat.' In this paper we report cognate studies in which we investigated the effects of the pyrrolidinylmethyl modification on antihypertensive activity in the SHR with a series of benzazepinone and benzothiazepinoneanalogs 5. Our data indicate that incorporation of the pyrrolidinylmethyl N-1 substituentgenerally increases duration and/or potency of antihypertensive activity in SHR without a large or consistent effect on potency in vitro. We also present data indicating that compounds possessing a substituent on the fused aromatic ring are more resistant to metabolic deacylation of the C3 hydroxy function which, in some cases, also results in a significant attenuation in calcium channel blocking activity. These observations appear to explain why aromatic substituents also frequently increase antihypertensive potency and/or duration

in the absence of a profound effect on potency in vitro. Taken together, our data indicate general structural modifications in both the benzazepinone and benzothiazepinone series that potentially improve antihypertensive activity as a direct result of increasing oral bioavailability. Synthesis

Compounds 1 were prepared by previously published meth~ds.l*~ Nonracemic If (see Table I) was prepared by simple hydrogenolysis (Pd/C, H2) of the corresponding nonracemic 7-chlorobenzazepinone l h (see ref 91.l (4) Floyd, D. M.; Moquin, R. V.; Atwal, K. 5.;Ahmed, S. 2.; Spergel, 5. H.; Gougoutas, J. 2.;Malley, M.F. Synthegie of Benzazepinoneand 3-MethylbenzothiazepinoneAdcgueg of Diltiazem. J. Org. Chem. 1990,55,5672-5579.

Das et al.

2612 Journal of Medicinal Chemistry, 1992, Vol. 35, No. 14

The alkylating agents 7a or 7b used to prepare compounds 5 and 10 in Table I were prepared from commercially available (S)-pyrrolidine-2-methanol by initial reaction with di-tert-butyldicarbonate(BOC20)in methylene chloride giving BOC-2-pyrrolidinemethanolfollowed by treatment with either p-toluenesulfonyl chloride in pyridine to form the tosylate 7a or with carbon tetrabromide and triphenylphosphine in ether to afford the bromide 7a or with carbon tetrabromide and triphenylphosphine in ether to afford the bromide 7b (see Scheme IIL5 Compounds 5 and 10 are single enantiomers; 3R,4S (benzazepinone numbering) for X = CH2 and 2S,3S (benzothiazepinone numbering) for X = S in Table I. Benzazepinones 5/10a and 5/lOc were prepared from racemic 6 whereas 5/10b and the benzothiazepinones 5/1Od and 5/1Oe were prepared from the nonracemic 6. Benzazepinones 6a-c and the benzothiazepinone 6e were prepared by previously described methods.'~~For example, alkylation of racemic 6a with bromide 7b in DMF at 80 "C formed a diastereomeric mixture of adducts which were separated by silica gel column chromatography to give the desired diastereomer 8a in 28% yield (56% of theory). Alternately, nonracemic 6d (see below for preparation) was alkylated with tusylate 7a in presence of cesium carbonate in DMF at 50 "C to afford 8d in 76% isolated yield after silica gel chromatography. In general, removal of the BOC group of 8 in dichloromethanewith Wluoroacetic acid and subsequent treatment with anhydrous hydrogen chloride afforded 10,as hydrogen chloride salts,in excellent yields. The 3-acetoxy derivatives 5 were prepared by initial treatment of 8 with acetic anhydride in dichloromethane in the presence of 4-(dimethy1amino)pyridineto form 9. These compounds were then deprotected and converted to their hydrochloride salts, as described above, to afford 5 in good yields. A third method was employed in the m e of 5/10c. For these compounds, separation of the diastereomeric mixture of 8c was not synthetically useful. Consequently, the mixture of diastereomers was carried through the sequence shown in Scheme I1 with final purification effected by recrystallization of 5c. The corresponding C3 hydroxy compound 1Oc was obtained by simple basic hydrolysis of 5c. The absolute stereochemicalassignments for compounds 5/10a and 5/lOc,which were prepared from racemic 6a and 6c,were assumed to be 3R,4S (benzazepinone numbering). This was based on both the sign of the optical rotation, which does not change with the additional stereocenter contained in the pyrrolidinylmethyl group, and their potent biological activity. The enantiomeric 3S,4R benzazepinones are known to possess very low potency in vitro.' The assignment of absolute stereochemistry for 5c was supported by single crystal X-ray analysk6 The absolute stereochemical purity of 5/10a and 5/10c was predicated on the observation that the synthetic precursors, 8a and 8c respectively, were chromatographically homogeneous. As noted previously, relative stereochemical purity is easily determined by NMR method^.^ Laborious methods have been published for the preparation of nonracemic benzothiazepinone 6d involving the separation of diastereomers formed by the acylation of (5) For a related preparation of 7a see: Jones, K.; Woo,K-C. A Total Synthesis of (-)-Ruepolinone. Tetrahedron 1991, 47, 7179-7184. (6) Gougoutes, J. 2.;Malley, M. F. Unpublished results. Both the relative and absolute stereochemical assignments for Sc and bd were confiied by single crystal X-ray analysie. The absolute stereochemical assignmentswere based on the known absolute configuration of 7a,b.

Scheme 1II.O Synthesis of 6d

11

12

Racemic

D-(-)-Tcu& Acid Salt 2S,3S

13

6d 2S,3S

a (a) &(-)-Tartaric acid, 95% EtOH; (b) NaOH, MeOH-H20; (c) eth~1-[3-(dimethylamino)propyl]carbodiimide, pH 4.5.

racemic 6d.7 As described in Scheme 111, we have developed a much more efficient method based on resolution at an early stage of the synthesis. This was possible due to the fact that the aromatic amino group of 11 was sufficiently basic to form stable salts with a variety of chiral carboxylic acids, presumably due to the presence of the p-methoxy group. The intermediate 11 was prepared in 6390yield by the reaction of 2-amjn&-methoxythiophenol with the requisite methyl p-methoxyphenylglycidate in refluxing toluene as described in the patent literaturee7 Treatment of the racemate 11 with D-(-)-t"iC acid in 95% ethanol gave, after a single recrystallization, 12 as its tartaric acid salt in 83% of the theoretical yield. This material was then hydrolyzed to the carboxylic acid 13 and cyclized to 6d via a carbodiimide-mediated cyclization in aqueous DMF in an overall yield of 96%. None of the enantiomer was detected by NMR by employing chiral shift reagents. Further proof of absolute configurationwas obtained from the X-ray crystallographic structure determination of 5d which was prepared from 6de6 Results and Discussion The data presented in Table I1 indicate that for several of the compounds incorporation of the pyrrolidinylmethyl group improves antihypertensive activity in the SHR following oral administration. Furthermore, the effect on antihypertensive potency and/or duration is not consistent with any increase in potency in vitro as a result of this structural modification. For example, the largest increasea in potency in vitro associated with the pyrrolidinylmethyl N1 modification are with the chloro-(la vs sa) and methoxy-(lc vs 5c) substituted benzazepinones, 7- and Idfold, respectively! However, despite being 7-fold more potent, Sa is only slightly more active as an antihypertensive agent than its (NJV-dimethy1amino)ethylanalog la. With the methoxy-substituted benzazepinones, the pyrrolidinylmethyl derivative 5c demonstrates markedly increased antihypertensive activity compared to IC, although it is only about 4-fold more potent in vitro. In (7) Takeda, M.; Ohiahi,T.; Nakajima, H.;Nagao, T. U.S.Patent 4,590,188. (8) Compounds poeeessing the wrong absolute stereochemistry are at leaat 10 times less potent than the biologically most active enantiomers. It is, therefore, an acceptable working assumption that (3R,4S)-benzazepinones and (2S,3S)-benzothiazepinones are twice aa potent in vitro aa their racemates; ~ c ~assumption ~. is that is, ICwwd) = 0.5 X ( I C s o ~ , , o ~This supported by our observationsm thle series of compounds (see refs 1 and 3).

Journal of Medicinal Chemistry, 1992, Vol. 35, No.14 2613

Benzazepinone Calcium Channel Blockers

Table 11. Vasorelaxant and Antihypertensive Activity of Compounds 1, 5, and 10 .OMe

OMo

5

1

10 ~

% decrease in BP ca. 135 pmollkg PO

compd' lad Sa 10a lb Sb 10b 1Cd

R1

Rz

X

ICw (pMlb 04h 6-12 h 12-18 h ne c1 H CH2 1.60 (0.71-3.40) 10 (*4) 8 (*5) 11 (*3) 5 17 (tl) c1 H CH2 0.12 (0.08-0.17) 20 (14) 6 l8 (14) (*3) 15 (h4) 0.13 (0.09-0.20) 19 13 ( t 4 ) 5 c1 H CH2 0.15 (0.11-0.20) 29 (*3) 29 (k2) 29 (13) 10 H CH2 CF3 38 (*2) 38 (*2) 6 0.09 (0.07-0.12) 44 (*2) H CH2 CFS 0.10 (0.07-0.13) 32 (*3) 21 (*1) 21 (*4) 6 H CH2 CFS OMe CH2 1.90 (0.88-4.00) 15 (13) 8 (14) 10 (13) 5 H sc H OMe CH2 0.23 (0.15-0.36) 39 (*2) 33 (tl) 31 (*2) 5 1OC H OMe CH2 1.12 (0.68-1.84) 18 (*3) 19 (*3) 18 (14) 5 Idd H OMe S 0.13 (0.11-0.20) 34 (*3) 21 (*6) 19 (16) 5 Sd H OMe S 0.12 (0.08-0.19) 42 (+2) 39 (*3) 37 (13) 6 1Od H OMe S 0.37 (0.23-0.60) 32 (*3) 33 ( t 4 ) 23 (*3) 6 le H H S 0.21 (0.13-0.36) 20 (*7) 12 (*8) 19 (16) 10 8 (a31 10 ( t 3 ) 9 se H H S 0.34 (0.23-0.60) l8 10 (*2) 10 (*2) (*3) 9 (*2) 100 H H S 1.32 (0.79-2.22) 8 13 (*4)e 5 8 (*3) 8 (*2) If H H CH2 2.99 (1.69-5.30) "All compounds are the biologically most active enantiomers unless noted otherwise;2S,3S for X = S and 3R,4S for X = CH2. bAverage value from at least four experiments;95% confidence range shown in parentheses. 'Data reported as % fall in blood pressure (A standard error) from control levels where n = the number of SHR from which data was recorded. Racemic compound. e SHR data from racemic compound dosed at 270 pmol/kg po.

other examples, significant increases in antihypertensive activity are observed with no effects on potency in vitro, i.e., l b vs 5b and Id VB 5d. Consequently, we ascribe the increase in antihypertensive potency and/or duration of action observed with pyrrolidinylmethyl-containing compounds as being associated with increased metabolic stability compared to their (Nfl-dimethylamho)ethyl analogs and not to an effect on intrinsic calcium channel blocking activity. This assumption is supported by the results obtained from metabolism studies with the trifluoromethyl-substituted benzazepinones l b and 5b.' For the methoxy-substituted benzazepinone and benzothiazepinone analogs, incorporation of the pyrrolidinylmethyl N1 substituent resulted in compounds (5c,5d) that possessed potency and duration of action similar to the trifluoromethyl-substitutedbenzazepinone 5b. The benzazepinone IC demonstrates a small antihypertensive effect which essentially dissappears &r 6 h. By comparison, the pyrrolidinylmethyl analog 5c shows marked antihypertensive activity throughout the 18 h test. The effect of the pyrrolidinylmethyl modification in the methoxy-substituted benzothiazepinone series becomes clearer on comparing the apparent duration of action of ld and 5d. Although Id is a potent antihypertensiveagent in this test, the magnitude of ita antihypertensive effect decreased by approximately 50% over the duration of the test in SHR. In contrast, 5d, which incorporates the pyrrolidinylmethyl N1 functionality, maintains an antihypertensive effect that is essentially unchanged over 18 h. Thus, the increased metabolic stability of the pyrrolidinylmethyl group compared to the (NJV-dimethylamino)ethyl moiety can manifest itself aa an effect on the duration of action and/or the potency of the compound under study. With the chloro-substituted benzazepinones l a and 5a, the effect is much less pronounced? Furthermore, despite

the fact that the chlorobenzazepinone 5a is equipotent in vitro compared to the benzazepinones 5b,c and benzothiazepinone 5d, it is much less active in the SHR. Similarly, the benzothiazepinone 58, the pyrrolidinylmethyl analog of diltiazem (le),shows reduced antihypertensive activity compared to 5 H , despite similar potency in vitro. Although there are many factors that affect oral activity and that could explain these results, based on the studies described below, we believe 6a and 5e possess lower antihypertensive activity as a result of additional metabolic processes. Our overall conclusion from these studies is that incorporation of the pyrrolidinylmethyl N1 substituent is a rational approach to generating compounds with increased oral antihypertensive activity as a result of the increased resistance to oxidative metabolism of this functionality compared to the (Na-dimethy1amino)ethyl ~

~~

(9) In general, we believe the comparison of antihypertensiveactivity in SHR between racemic and nonracemic compounds is valid for this series of compounds. We find the theoretical 2-fold increase in potency in vitro between nonracemic and racemic compounds does not have a marked effect on antihypertensive activity. Thia is based on data from ref 1 ahown in the following table (for compound 1, X = CHJ. %

Decrease in BP

@ 135 pmollkg p.0.

ComDound EI

li Ij

Ez J b U h l i U

UOLuM1

U~~

H H

CI CI

Ac Racemic Ac 3R.4S

1.5 (0.84-2.8) 0.82 (0.54-1.3)

13 12

14 12

12

H

CF3 CF3

Ac M c Af 3R,4S

1.9 (0.95-3.6) 1.6 (0.87-3.1)

19 19

19 22

18 19

0.24 (0.13-0.45) 0.15 (0.11-0.20)

24 31

26 27

24 30

H

lk Ac M c lb Ac 3R,4S a. Dosed at 270 WmoVkg p.0.

I8a

2614 Journal of Medicinal Chemistry, 1992, Vol. 35, No.14

N1 substituent. With analogs containing the latter substituent, oxidative deamination is a significant metabolic process and results in metabolites devoid of calcium channel blocking activity (see Scheme 11.l The notable exceptions to the generalization stated above are with diltiazem (le) and the chloro-substituted benzazepinone la. In these examples, incorporation of the pyrrolidmylmethyl group did not result in an increase in antihypertensiveactivity in the SHR. As noted above, one hypothesis to explain this observation is metabolic deacylation. In the case of diltiazem (le and 58, loss of the C3 acetyl group significantly attenuates potency in vitro.*O However, this hypothesis is not generally supported by examination of the data for other C3 hydroxyl analogs (10) shown in Table 11. The data from the compounds in Table 11that demonstrate an increase in antihypertensive potency and/or duration of action with the incorporation of the pyrrolidinylmethyl group give little insight into the importance of metabolic deacylation. For (trifluoromethy1)benzazepinone Sb, loss of the C3 acetyl group does not attenuate potency in vitzo. Further, the C3 hydroxyl analog lob demonstrates significant, although somewhat reduced, antihypertensive activity relative of Sb. With the methoxybenzothiazepinone5d, the desacetyl derivative 1Od is only 3-fold leas potent in vitro and retains a level of antihypertensive activity relative to Sd similar to that observed with the trifluoromethyl analogs. The largest effect on potency in vitro resulting from C3 deacylation is observed for the methoxybenazapinone 5c. In this case, the desacetyl derivative 1Oc is &fold less potent than the parent 5c. However, like the examples above, 1Oc retains significant antihypertensive activity. Thus, none of these examples supply clear data regarding the potential for metabolic deacylation to attenuate the antihypertensive activity of the C3 acetyl derivatives 5. In the examples where there is a clear failure of the pyrrolidinylmethyl group to increase antihypertensive potency compared to the corresponding (NJV-dimethylaminolethy1 analogs, the data can be interpreted to indicate a role for additional metabolic pathways. For example, metabolic deacylation of Sa would give the 3hydroxy derivative loa. In this case, the experimental results could be explained by the rapid metabolic conversion of Sa to loa since both compounds demonstrate the same moderate level of antihypertensive activity. It is important to note, however, that both compounds possess identical potency in vitro and the lack of potent antihypertensiveactivity could be due to a variety of other factors including inherently poor oral absorption. The effect on potency in vitro from loss of the C3 acetyl group from 58 in the diltiazem (le) series is more notable. In this case the desacetyl derivative 1Oe shows a 4-fold attenuation of vasorelaxant potency. However, as in the above example, both 58 and 1Oe lack potent antihypertensive activity which could be interpreted to indicate that upon oral dosing, 58 is rapidly converted to 1Oe. Thus, in both of these examples, the failure of the pyrrolidinylmethyl analogs Sa,e to demonstrate improved antihypertensive activity over their (NJV-dimethy1amino)ethylderivatives 1a.e could be explained by rapid metabolism to the C3 hydroxy materials 10a,e which show reduced antihypertensive activity, (10) Schoemaker, H.;Hicks, P. E.;Langer, S. Z. Calcium Channel Receptor Bind- Studies for Diltiazem and Its Major Metabolites Functional Correlation to Inhibition of Portal Vein Myogenic Activity, J. Cardbvasc. Phurmacol. 1987,9,173-180. See elso ref 2.

Doe et 41. Table 111. Hydrolysis of C3 Acetyl Croup by Rat Liver Homogenate extent of hydrolysis - - ( . .)aI compd Ri R2 x 5minb 30min le H H S 68 100 be H H S 32 57 CH2 18 53 lf H H CHZ 7 17 Sa H C1 Sd H OMe S 3 7 CH2 3 3 bo H OMe CH2 0 2 lb CFS H CH2 0 0 Sb CFB H a (Amount of 3-hydroxy product/total amount of substrate-related compound) x 100. Incubation period.

In an attempt to explain the apparent discrepancies in our data we investigated the ability of rat liver homogenate to hydrolyze the C3 acetyl group of several compounds in Table 11. The data from this study are summarized in Table I11 and indicate that the presence of an aromatic Substituent generally inhibits hydrolysis of the C3 acetyl group by the liver homogenate esterases. Comparison of the extent of hydrolysis of le and If at 5 and 30 min indicates that the benzazepinone system may be hydrolyzed somewhat more slowly compared to an analogous benzothiazepinone. Nevertheless, both compounds are significantly hydrolyzed after 30 min. Similarly, independent of the class of compounds, neither the methoxybemthkepinone 6d nor the analogous benzazepinone Sc is extensively hydrolyzed, indicating that the methoxy substituent inhibits deacylation under these conditions. The data also demonstrate that the nature of the N1 substituent does not affect the ability of these compounds to act as substrates for the liver homogenate esterases. Both of the benzothiazepinones le and 58, which differ only in the structure of the N1 substituent, are significantly hydrolyzed under these conditions. Also, independent of the N1 substituent, the C3 acetyl groups of the (trifluoromethylbnzazepinones Ib and Sb are resistant to esterase activity. From Table 111,it appears that the most important factor determining susceptibility to enzymatic deacylation is the presence or absence of an aromatic substituent, with the degree of protection being possibly dependent on the nature of the substituent (compare the data for the chlorobenzazepinone Sa to those of Sb or 5c in Table 111). The results of this study appear to explain some of the discrepancies in Table 11. These data support the hypothesis that the unsubstituted benzothiazepinone 58 shows no increase in antihypertensive activity compared to its (NJWdimethy1amino)ethylanalog le as a result of extensive metabolism to the less active 3-hydroxy derivative 1Oe. Conversely, the methoxy-substituted benzazepinone Sc demonstrates a high level of antihypertensive activity because it is resistant to metabolic deacylation to the 3-hydroxy analog 1Oc. The studies summarized in Table IIIare also consistent with our previous obaervations that replacement of the C3 acetyl group of benzazepinone If with alkyl substituents that are resistant to metabolism results in analogs with good antihypertensive activity in the SHR.3J1 In this case, C3 metabolism to a less active (11) Dae, J.; Floyd, D. M.; Kimball, S. D.; Duff,K. J.; TNCChi Vu; Lago, M. W.;Moquin, R. V.; Gougouh, J. Z.; Malley, M. F.;

Moreland, S.; Brittain, R. J.; Hedberg, S. A.; Cucinotta, G. G. Benzazepinone Calcium Channel Blacking Agenta. 3. Synthesis and Structure-Activity Studies of 3-Alkyl-Benzazepinones. J. Med. Chem. 1992,36,773-780.

Benzazepinone Calcium Channel Blockers

Journal of Medicinal Chemistry, 1992, Vol. 35, No.14 2616

analog is precluded by the alkyl group, and the presence of an aromatic substituent is not necessary for the expression of antihypertensive activity in SHR. This study also appeara to explain why simple methoxylation of the diltiazem nucleus (compare Id to le), while having no significant effect on intrinsic potency, resulta in a marked increase in antihypertensive activity. A similar effect is noted by comparing diltiazem to ita 8-chloro analog TA 3090.1 The data in Table 111 may also explain the weaker antihypertensive activity of chlorobenzazepinoneSa relative to the other aromatiesubstituted derivatives in Table 11. Although the chloro substituent apparently affords some protection against enzymatic deacylation compared to compounds lacking an aromatic substituent, Sa is still deacylated significantlymore than either the methoxy- or Muoromethyl-substituted benzazepinones. These observations are consistent with the hypothesis that Sa is metabolized to the C3 hydroxy derivative 10a and may explain why, despite similar potency in vitro to the other aromatic substituted analogs 6 in Table 11, Sa produces a lower antihypertensive effect in SHR. Thus, based on this observation and the fact that Sa and loa possess identical antihypertensive activities when orally adminiatered to SHR,we conclude that the metabolic deacylation of Sa attenuates ita antihypertensive activity also. In summary, the studies described above illustrate that increased antihypertensive activity is obtained with diltiazem-like calcium channel blockers by limiting metabolic degradation of the critical basic amino functionality in the N1 substituent. We have shown that incorporation of the pyrrolindinylmethylN1 substituent is consistentwiht both significantly increased metabolic stability of the basic amino group and maintenance of high biological activity and resulta in compounds demonstrating potent and long-lasting antihypertensive activity in the SHR.' In addition, for compounds containing a C3 acetoxy substituent, maximal antihypertensive activity also requires the presence of a substituent on the fused aromatic ring. Although we can not rationalize this observation on mechanistic grounds, the aromatic substituent appears to inhibit metabolic deacetylation of the C3 acetoxy group and,consequently, the formation of 3-hydroxy metabolites which are frequently less active antihypertensive agents than their corresponding 3-acetoxy analogs. The degree of this effect is, however, dependent on the nature of the substituent. We believe the combined resulta from these studies indicate general methods to improve the pharmacokinetic profile, and hence, the oral antihypertensive activity, of diltiazem-like calcium channel blocking agents. More generally, the approach we have taken to increasing the potency in vivo of diltiazem-like compounds by having our synthetic effort driven in part by an understanding of important metabolic pathways would be applicable to a variety of medicinal chemistry programs.

of the test compounds. All blood pressure measurements were obtained by direct recording. A more detailed description of theae biological test methods hae been reported previously.' Rat Liver Homogenate Deacylations. Enzymatic hydrolysis of the C3 acetyl functionalitywas performed by exposing the teat compound (0.4mM final concentration) to rat liver homogenate (2mL, 0.3 g of liver/mL) in the presence of phoephate buffer (5 pM, f i a l concentration, pH 7.4) in a total volume of 2.5 mL at 37 OC. The mixture was incubated for either 5 or 30 min and diluted with acetonitrile (2.5 mL), centrifuged to remove precipitated protein, and analyzed for the expected hydrolysis product by HPLC with an isocratic mobile phaaq (50mM ammonium 4 0 by volume; flow rate 1 acetate buffer, pH hcetonitrile, a mL/min). HPLC analysis was performed using a Beckman Ultrasphere-ODS column (5 m, 4.6 X 250 mm; Beckman Instruments, Inc., San Ramon, CA) on a Perkin-Elmer Seriee 410 pump system, equipped with an LC-235diode array detector (Perkin-Elmer Corp., Norwalk, CT) set at 230 nm. General Synthetic Procedures.Melting pointa were recorded on a Thomas-Hoover capillary apparatus and are reported uncorrected. Proton NMR ('H NMR) spectra were obtained on JEOL FX-270or GX-400 spectrometersand are reported relative to tetramethylailane (TMS) reference. Carbon NMR ('9c NMR) data were obtained on the JEOL FX-270 or FX-6OQspectrometers and are also reported relative to TMS. Analysis by TLC was performed on E. Merck Kieaelgel60, F-254silica-gel platea. Flash chromatography was done on either Whatman LPS-1 or E. Merck Kieselgel60 (240-400 mesh) silica gel. Unless otherwise stated, concentrationsof solutionswere done in vacuo. All reactions were conducted under an atmosphere of argon employing reagent grade solvents. The DMF was dried by storage over %A molecular sieves. Anhydrous THF was prepared by distillation from sodium benzophenone prior to use. (S)-[N-[ (tert-Butyloxy)carbonyl]-2-~yrro~~yl]methyl p-Toluenesulfonate (7a). A solution of (8-2-pyrrolidinemethanol (25g, 247 mmol) and di-tert-butyl dicarbonate (64.7 g, 297 mmol) in dichloromethane (700mL) was stirred at room temperature for 5 h. Removal of solvent afforded the desired BOC-proteded intermediate (49.7g) as an oil. A solution of this crude product (20.6 g, 102.4 mmol) in pyridine (100 mL) was treated with p-tolueneaulfonyl chloride (23.4g, 122.8 "01) and the resulting solution was stirred at room temperature for 5 h. Additional p-toluenesulfonyl chloride (9.8 g, 51.2 mmol) was added. After 23 h, the mixture was diluted with EtOAc and washed with saturated CuSO, solution (3X).The EtOAc extract was dried (MgSOI), filtered, and concentrated. The crude oil was chromatographedon a silica gel column and eluted with 10-3046 EtOAc in hexanes to obtain 7a (32.1g, 88% overall yield) as a viscous oil: ["ID = -ao(c = 1, methylene chloride), lit." ["ID -39.6O (c = 0.74,methylene chloride); 'H NMR (CDCls) 6 7.80 (d, 2 H),7.25 (d, 2 H),4.10 (m, 1 H), 3.90 (bs, 2 H), 3.30 (m, 2 H), 2.45 (8, 3 H), 1.70-1.95 (m, 4 H), 1.40 (bs,9 H). (S)-[N-[(tert-Butyloxy)carbonyl]-2-~y~~~yl]methyl Bromide (7b). A solution of crude (5')-N-[(tert-butyloxy)carbonyl]-2-pyrrolidinemethano1(49.7g), carbon tetrabromide (129.5g, 494 "01) and triphenylphosphine (164g, 494 "01) in ether (1L) was stirred overnight at room temperature. The precipitated solids were then filtered and washed with ether. The filtrate was concentrated to obtain a yellow oil which was chromatographed on a silica gel column. Elution with hexanes, followed by 30% EtOAc in hexanes gave 7b (36.99g, 57% overall yield) as a colorless oil: 'H NMR (CDClJ 6 4.00 (m, 1 H), 3.36 (m, 1 H), 3.37 (m, 3 H), 2.1-1.85 (m, 4 H),1.47 (8, 9 H).This material was used without further characterization. (3R,4S)-gChloro-lr,4p-tetrahydro-3-hdro~-k( kmeth-

Experimental Section Biological Testing. To determine the effects of structural modifications in the bemazepinone and benzothiazepinone seriea on potency in vitro we have measured the vasorelaxant effects of compoundsin circumferentialstrips of potassium-depolarized rabbit aorta The ICw value reported represents the concentration of compound necessary to cause 50% relaxation of a maximal contraction in reaponw to 100 mM KCL Antihypertensiveactivity was calculated as the percent fall in systolic blood pressure from p " g control value and is reported as the mean value acquired from five rats at a standard dose of 135 pmol/kg PO unless otherwise noted. The maximum antihypertensive effect is reported for the W h , 6-124, and 12-18-htime periods after oral doeing to give an indication of both the potency and duration of action

-

oxyphenyl)-l-[[N - [ ( tert -butyloxy)carbonyl]-2pyrrolidlnyl]methyl]-2H-l-benzazepin-2-one(Sa). Method I. Solid (cis)-gChloro-l,3,4,6tetrahydro-3-hydroxy-4-(4metho ~ ~ n y l ) - W - l - b p i n - 2 - o (6a, n e 3 g, 9.44m o l ) was added to a suspension of NaH (270mg, 11.33 mmol) in DMF (95mL). After 1 h, 7b (3g, 11.33 mmol) was added and the mixture was heated to 80 OC for 3 h. Additional NaH (110mg, 4.58 mmol) were added, and heating was continued and 7b (1.25g, 4.72 "01) for another 2 h. The mixture was cooled, diluted with water, and extracted with EmAc (3X). The EtOAc extracts were combined, washed with 10% aqueous LiCl solution (3X),dried (MgSO,),

Das et al.

2616 Journal of Medicinal Chemistry, 1992, Vol. 35,No. 14

54.7,50.6,45.5,28.9,24.2,20.9. Anal. (cuH&&jS*Hc1) C, H, N, C1, S. [3R-[ 1 5*,3~,~]]-3(Acetyloxy)-1,3,4&tetmhydr0-7-methoxy-4-(4-methoxypheny1)-1 (2-pyrrolidinylmethyl)-2H-1 lmnzazepin-2-one, Monohydrochloride (Sc). Method 111. A solution of cis-l,3,4,5-tetrahydro-3-hydroxy-7-methoxy-4-(4methoxyphenyl)-W-l-benzazepin-2-one (see ref 1 for method of preparation) (6c; 3 g, 9.57 mmol), 7a (6.8g, 19.13 mmol), and cesium arbonate (6.24g, 19.15mmol) in DMF (30 mL) was heated to 50 OC overnight. Extractive workup, following the procedure d d b e d in preparation of 8d,and subsequent chromatography on a silica gel column with 33% EtOAc in hexanes as eluent afforded a 1:l mixture of & and ita diastereomer (4.58g, 96%) as a foam. A solution of this mixture (4.55g, 9.16 mmol), acetic anhydride (4.32mL, 45.79 mmol), and 4-(dimethylamino)pyridine (2.24g, 18.34"01) in dichloromethane (142mL) was stirred at room temperature for 24 h. The reaction was concentrated, and [25-[2a,3u,S(R *)I]-6-[[1-[(1,1-Dimethy1methoxy)carbonyl]-2-pyrrolidinyl]methyl]-2,3-dihyd~3-hydrox~8- the residual oil was chromatographedon a silica gel column (50% EtOAc in hexanes) to give a mixture of 9c and ita diastereomer methoxy-2-(4-methoxyphenyl)- 1,S-benzazothiampin-4(4.05g, 82%) as a yellow foam. A solution of 9c and ita diaste(SH)-one(Sa). Method 11. A solution of (2S,3S)-2,3-dihydro3-hydroxy-8methoxy-2-)4-methoxyphenyl)-l,bbenzothiampin- reomer (4.25 g, 7.89 mmol) and trifluoroacetic acid (36mL) in dichloromethane (36 mL) was stirred at room temperature for 4(%ne (Sa, 1 g, 3.02 mmol), 7a (340mg,0.97 mmol), and cesium 1 h. The reaction was then concentrated and the residual oil carbonate (2.15g, 6.05 "01) in DMF (10mL) was heated to 50 bid in M A C(150mL) and waehed with "ted NaHcoB OC overnight. The mixture was cooled, diluted with water (10 solution (100mL, 2x1. The EtOAc extract was dried (MgSO,), mL), and extracted with ether (30mL, 3X). The ether extracts filtered,and concentrated by mevaporation with ether. The yellow were combined, washed with 10% aqueous LiCl solution (30 mL, solid waa cryatallid from acetonitrile to obtain enantiomerically 3X),dried (MgSOJ, liltered,and concentrated. The residual foam pure k (910mg,63% of theow mp 287-288 "c,[a]D +102.7" was chromatographedon a silica gel column and eluted with 50% (c = 1, MeOH); 'H NMR (CDC13)S 7.38 (d, 1 H), 7.16 (d, 2 H), EtOAc in hexanes to obtain 8d (1.19g, 76%): mp 75-80 "C; [a]~ 6.89 (m, 4 H), 5.17 (d, 1 HI, 4.48 (dd, 1 H), 4.27 (bs,1 H), 3.84 = +119" (c = 1, MeOH). (8, 3 H), 3.81 (8, 3 H), 3.52 (m, 1 H), 3.37 (m, 1 H), 2.89 (m, 2 H), [3R -[1S *,3a,k]]-6-Chloro-1,3,4,5-tetrahydr0-3-hydroxy2.22 (m, 2 H), 1.88 (s,3H); 13CNMR (CDC13)6 20.1, 22.9, 28.3, 4 44-methoxypheny1)-1-(2-pyrrolidinylmethyl)-2H-1-bemz37.5, 46.4, 49.9, 54.8, 55.2,58.8, 71.3, 113.5, 114.8,123.8, 128.8, azepin-2-0110(loa). A solution of 8a (670mg, 1.34 "01) and 131.4,133.9,134.6,158.2,158.5,169.6,170.1. Anal. (C2aH&J2in dichloromethane (10 trifluoroacetic acid (1 mL, 13.1 "01) 06.HCLO.25HzO) C, H, N, C1. mL) was stirred a t room temperature for 6 h. The mixture was t h -2-Hydroxy-3-[(2-amino-6-methoxyphenyl) thiol-3diluted with Saturated KHC03 solution and extracted with ethyl (4-methoxypheny1)propionic Acid, Methyl Ester (11). A acetate (3X). The ethyl acetate extract was dried (MgSOJ, fiisuspension of trans-3-(4-methoxyphenyl)oxiranec.arboxylicacid, tered, and concentrated. The crude residue was chromatographed methyl ester (50.5g, 243 m o l ) and 2 - a m i n ~ b m e t h o ~ o p h e n o l on silica gel plate and eluted with 10% MeOH in dichloromethane (42.9.g. 276 "01) in toluene (300mL) was refluxed for 6 h with (280mg,52%): mp 80-84 "c,[a]~ = +148" (C = to obtain a slow stream of nitragen pawing over the reaction. The mdting 1, MeOH); 'H NMR (CDC13) 6 7.33 (d, 1 H), 7.23 (t, 1 H), 7.21 solution was then cooled and concentratedto a yellow-brown solid (d, 2 H), 7.20 (dd, 1 H), 6.90 (d, 2 H), 4.46 (dd, 1 H), 4.18 (t, 1 which was recrystallized from ethanol twice to yield the desired H), 3.81 (8, 3 H), 3.65 (m, 2 H), 3.50 (m, 2 H), 2.99 (t, 1 H), 2.90 adduct 11 (52.7g, 63% yield): mp 101-103 "C (lit? 99.5-102.5 (m, 2 H), 2.65 (bs, 2 H), 2.00-1.50 (m, 3 H); 13C NMR (CDC13) "C). Anal. (ClaH21NO5S) C, H, N, S. S 173.0,159.0,142.4,134.0,132.5,130.8,129.8,128.2, 127.8,121.2, (2535 )-2-Hydroxy-3-[(2-amino-5-methoxyphenyl)thio]113.9, 69.6, 58.0, 55.2, 52.5, 51.3, 46.5, 32.7, 29.9,25.2. Anal. 3-(4-methoxyphenyl)propionic Acid, Methyl Ester, D-(-)([email protected]) C, H, N, C1. Tartaric Acid Salt (12). A mixture of 11 (5.45g, 15 "01) and [~S-[~U&,S(R*)]]-~-(AC&~~OX~)-S-[[ 1-[( 1,l-dimethylethtarta tart arc acid (2.50g, 16.7 "01) was slurried in 95% ethanol oxy)carbonyl]-2-pyrrolidmyl]methyl]-2,3-dihydro-8-met hoxy-2-(4-methoxyphenyl)-l,S-benzazothiazepin-4(SH)-one (50 mL) and heated to effect solution. The warm solution was cooled in a water bath which resulted in the formation of a (9d). A solution of 8d (1.23g, 2.4 mmol), acetic anhydride (1.25 crystalline product. After standing for 3 h, the solid was filtered, g, 12.2 mmol), and 4-(dimethylamino)pyridine(640mg,5.2 m o l ) washed with cold 95% ethanol, and dried. This material (3.73 in dichloromethane(25mL) was stirred at room temperature for g) was recrystallid once from 95% ethanol (40mL) to afford 24 h. The solvent was removed under reduced pressure, and the 3.20 g (83%of theory) of the desired product: mp 168-169 "C; residual oil was chromatographedon a silica gel column. Elution D I : [ = +140° (c = 1, methanol). Subsequent recrystallizations With methylene chloridemethanol (60:l) afforded 9d (1.17g, dld not change either the rotation or the mp of this material. AnaL 89%): mp 65-70 "c;[a]D = +98.5" (c = 1, MeOH). [2S-[2u&,S(R*)]]-3-(Acetyloxy)-2,3-dihydro-8-methoxy(Ci&ziNO&C4&08) C, H, N, S. (25,35 )-2-Hydroxy-3-[(2-amino-6-methoxypheny1) thiol2-(4-methoxyphenyl)-S-(2-pyrrolidinylmethyl)-1,S-benzo3-(4-methoxyphenyl)prpionicAcid (13). A sample of 12 (8.0 thiazepin-4(5H)-one, Monohydrochloride (Sd). A solution g, 15.6 mmol) was added to a solution of sodium hydroxide (4.0 of 9d (1.13g, 2.03 mmol) in dichloromethane (10mL) and trig, 100 mmol) in water (80 mL) followed by the addition of fluoroacetic acid (10ml)was stirred at room temperature for 30 methanol (80mL) which resulted in the rapid formation of a clear min. The solution waa concentrated and the residual pale orange solution. After 1 h, the reaction waa diluted with water (120mL) oil was dissolved in EtOAc (30mL) and treated with saturated and cooled with stirring while 1 N HCl(80 mL) was slowly added. aqueous KHCOs. The EtOAc extract was dried (MgSOJ, filtered, After continued cooling for 2 h, the resulting heavy precipitate and concentrated. The pale yellow solid was dissolved in acewas fiitered and dried to afford 5.35 g of 13 (98%): mp 188-189 tonitrile (10mL) and treated with a 5.1 N solution of hydrogen "C dec; [a]D = +286" (c = 1, DMF). Anal. (C17HlBNO@)C, H, chloride in ethanol (0.40mL). The solution was diluted with ether N, S. (50 mL) to obtain a solid (870mg,80%) which was recrystallized (25,35)-2P-Dihydro-3-hydroxy-8-methoxy-2-( 4-methowfrom acetonitrile (150mL) to obtain Sd (780mg): mp 228-230 phenyl)-lP-benzothiazepin-4(SH)sne(6d). A solution of 13 OC, [a]D = +49" (c = 1, MeOH); 'H NMR (CDC13) 6 7.52 (d, 1 (5.25g, 10.2 "01) in DMF (150mL) was treated with a solution H), 7.34 (d, 2 H), 7.27 (d, 1 H), 7.10 (dd, 1 H), 6.90 (d, 2 H),5.18 of ethyl [3-(dimethylamino)propyl]carbodiimidehydrochloride (d, 1 HI, 5.03 (d, 1 H), 4.44 (dd, 1 H), 4.25 (dd, 1 H), 4.00 (m, 1 (3.40g, 17.8 "01) in water (20mL). The pH of the solution was H),3.85 (s,3H), 3.82 (8, 9 H),3.41 (m, 2 H), 2.25-2.17 (m, 6 H), maintained at 4.5to 5.0 by the dropwise addition of 1 N HCl (ca. 1.88 (8,3 H); 'W NMR (CDC13)6 170.5,170.2,160.4,159.3,137.2, 10 drops) during the course of the reaction which appeared to 131.0,129.3,126.4,126.0,121.0,118.0,114.4,71.7,59.6,56.3,55.7,

filtered,and concentrated. The residual oil was chromatographed on a silica gel column and eluted with 620% EtOAc in hexanes to obtain 8a (650mg, 28%) 88 an oik [a]~ = +135.2" (C = 1, MeOH); TLC Rf = 0.60 (50% EtOAc in hexanes); 'H NMR (CDC13) 6 7.33 (m, 3 H), 7.19 (d, 2 H), 6.91 (d, 2 H), 4.55 (m, 1 H), 4.21 (t, 1 H), 3.81 (8, 3 H), 3.75 (m, 1 H), 3.45 (m, 3 H), 3.10 (9, 1 H), 2.82 (m, 2 H), 2.47 (t, 1 H), 2.20-1.70 (m, 4 H), 1.44 (be, 9 H); "C NMR (CDCl3) 6 173.3, 158.9,140.5,130.5,129.5,120.9, 113.9,69.6,55.2, 54.9, 51.4, 32.8, 28.5, 28.4. A sample of the diastereomer of 8a (foam, 420 mg, 18%) was ale0 isolated ["ID = 40.5" (c = 1, MeOH); TLC R = 0.53 (50% EtOAc in hemes); 'H NMR (CDC13)6 7.40-7.15 (m, 5 H), 6.90 (d, 2 H), 4.40-4.25 (m, 2 H), 4.18(t, 1 H), 3.81 (8, 3 H), 3.70 (m, 1 H), 3.45 (m, 2 H), 3.15 (q,1 H), 2.80 (m, 2 H), 2.47 (t,1 H), 2.20-1.70 (m, 4 H), 1.44 (be, 9 H); "C NMR (CDC13) S 158.9,154.5, 140.5,132.0,129.7, 128.3, 127.8,121.2,113.9,69.6,56.3,55.2,51.1,32.9,28.5, 28.3.

-

-

J. Med. Chem. 1992,35,2617-2626

2617

be complete after c& 3 min (no further addition of HC1 required). mg of 10%Pd/C), and the mixture was shaken overnight under The mixture was stirred for an additional M) min and poured inta 30 psi of hydrogen gas at room temperature. The mixture was 800 mL of ice-water resulting in the formation of a precipitate. then filtered through Celite and the filtrate, plus washes, was The mixture was cooled in an ice bath for an additional 4 h and washed with saturated NaHCOs. The organic layer was dried over filtered, and the mli& were washed with additional water. After MgS04and concentzated. The midue was diseolved in ether and drying, 6d (4.87 g, 98%) was obtained as a colorless solid: mp treated with HC1-saturated ether. The resulting white precipitate 190-191 OC; [a]D = +95O (c = 1,methanol), [lit. mp 187-190 OC, was filtered and dried to yield 0.13 g (83%)of the deaired product: mp 63-65 OC; [ a ]=~+117.2O (c 1,MeOH); 'H N M R (CDBOD) [a]D + 9 9 O (C = 0.26, DMF')].' Anal. (Ci,HI,NO4S) C, H, N, S. 6 7.45 (m, 3 H), 7.35 (m, 1H), 7.25 (d, 2 H), 6.90 (d, 2 H), 5.05 (3R- c h )-3-(Acetylosy)-l-[2-(dimethylamino)ethyl]- (d, 1H), 4.30 (m, 2 H), 3.80 (m, 4 HI, 3.65 (m, 1 H), 3.45 (m, 1 1~,4Ptetrahydro-4-(4-metho~phenyl)-2E-l-benzazepin-2H), 3.00 (m, 8 H),1.85 (8, 3 H); '% NMR (CDsOD) 6 20.35,37.94, one, Monohydrochloride (10. A solution of (3R-cis)-3-(ace44.05,45.34,51.53,55.74,56.46,73.19,114.71,123.76,128.88,129.95, tylo~y)-7-chlor0-1[2-(dimethylamino)ethyl]-1,3,4,5-tetrahydr0130.67,130.96,132.63,135.51,141.38,160.41,171.07,171.79. AnaL 4-(4-methoxyphenyl)-W-l-benzazepin-2-onemonohydrdoride (Cp+H&J204.HC1.1.5HzO) C, H, N. (lh, 170 mg,0.36 m o l , see ref 1for preparation) was neutralized Acknowledgment. Analytical data were supplied by in ether by washing with with saturated NaHCOS. The ether the Bristol-Myers Squibb Analytical Research and Desolution of the free base was then concentrated and the residue velopment Department. redissolved in acetic acid (10 mL). Catalyst was then added (80

Studies on Antidiabetic Agents. 11.' Novel Thiazolidinedione Derivatives as Potent Hypoglycemic and Hypolipidemic Agents2 Takashi Sohda,* Katautoshi Mizuno, Yu Momose, Hitoshi Ikeda, Takeshi Fujita, and Kanji Meguro Research and Development Division, Takeda Chemical Industries, Ltd., 17-85, Jusohonmachi 2-chome, Yodogawa- ku, Osaka 532, Japan. Received January 13,1992

In the course of further chemical modification of the novel antidiabetic pioglitazone (AD-4833, U-72,107), a series of 5[4(2-or 4azolybdkoxy)benzyl- or - b e n z y l i d e n e ] - 2 , 4 t o n e a was prepared and evaluated for hypoglycemic and hypolipidemic activities in insulin-resistant, genetically obese, and diabetic KKAY mice. Replacement of the 2-pyridyl moiety of pioglitamne by a 2- or 4oxazolyl or a 2- or 4-thiazolyl moiety greatly enhanced in vivo potency. The corresponding &benzylidene-type compounds, in which a methine was used as a linker between the benzene ring and the thiazolidmedione ring, also had potent biological activity. Among the compounds synthesized, 5[4-[2-(5-methyl-2-phenyl-4-oxazolyl)ethoxy]benzyl]-2,4-thiazolidinedione (18)exhibited the most potent activity, more than 100 times that of pioglitazone. The synthesis and structure-activity relationships for this novel series of derivatives are detailed.

Introduction

Insulin resistance is a characteristic feature of noninsulin-dependent diabetes mellitus (NIDDM), in particular when it is associated with obesity. This insulin-resistant state at the peripheral tissue level causes impaired glucose utilization leading to hyperglycemia3 Therefore, amelioration of insulin resistance with a drug would provide a novel and useful means of treating NIDDM. With regard to this concept3s4few drugs have, however, been (1) Part X Momose, Y.;Meguro, K.; Ikeda,H.; Hatanaka, C.; Oi,

S.; Sohda, T. Studies on Antidiabetic Agents. X. Synthesis and Biological Activities of Pioglitazone and Related Compounds. Chem. Pharm. Bull. 1991,39,1440-1445. (2) This work has been presented at the 1st Annual Meeting of Division of Medicinal Chemistry, The Pharmaceutical Society of Japan (December 4-6, Okayama in Japan, 1991; Abstract pp 62-63). (a) Meguro, K.; Fujita, T. Thiazolidine Derivatives, Their Production and Use. European Patent 177353,1986. (b) Meguro, K.; Fujita, T. Thiazolidine Derivatives, Their Production and Use. U.S. Patent 4 775 687, 1988. (3) (a) Reaven, G. M. Insulin-Independent Diabetes Mellitus: Metabolic Char~ristica.Metabolism 1980,29,445-454. (b) DeFronzo, R. A.; Ferrannini, E.;Koivisto, V. New Concepts in the Pathogenesis and Treatment of Noninsulin-Dependent Diabetes Mellitus. Am. J. Med. 1983,74 (Suppl. lA), 52-81. (c) Olefeky, J. M.; Kolterman, 0.G. Mechanisms of Insulin Resistance in Obesity and NoninsuLin-Dependent (Type 11) Diabetes. Am. J. Med. 1981, 70,151-168. (4) Reaven, G. M. Therapeutic Approaches to Reducing Insulin Resistance in Patients with Noninsulin-Dependent Diabetes Mellitus. Am. J. Med. 1983, 74 (Suppl. lA), 109-112.

studied, and exercise and calorimetric restriction are still the fundamental modes of treating NIDDM patients. It has been reported that sulfonylureas,the most commonly used oral hypoglycemics, potentiate insulin action in peripheral tissues by increasing the number of insulin re~eptors;~ however, their main mechanism of action actually involvea insulin secretion. This insulin secretion can bring about undesired effects such as induction of hypoglycemia. In 1982, we reported a series of 5-(4-alkoxybenzyl)-2,4thiazolidinedioness" as novel antidiabetic agenta which were shown to effectively reduce insulin resistance or potentiate insulin action in genetically diabetic and/or obese animals. Ciglitamne,a prototypical compound of the seriea (Chart I), was shown to normalize hyperglycemia, hyper(5) Gavin, J. R.,111Dual Actions of Sulfonylureas and Glyburide. Receptor and Post-Receptor Effects. Am. J. Med. 1985, 79 (Suppl. 3B), 34-42. (6) (a) Sohda, T.; Mizuno, K.; Imamiya, E.; Sugiyama, Y.;Fujita, T.; Kawamatsu, Y.Studies on Antidiabetic Agents. 11. Synthesis of 5444 l-Methylcyclohexylmethoxy)benzyl]thiazolidine-2.4dione (ADD-3878) and Ita Derivatives. Chem. Pharm. Bull. 1982,30, 3580-3600. (b) Fujita, T.; Sugiyama, Y.;Taketomi, S.; Sohda, T.; Kawamatsu, Y.;Iwatsuka, H.; Suzuoki, 2.Reduction of Insulin Resistance in Obese and/or Diabetic Animals by 5- [4l-Methylcyclohexylmethoxy)benzyl]thiazo( lidine-2,4-dione (ADD-3878, U-63,287, Ciglitazone), a New Antidiabetic Agent. Diabetes 1983, 32, 804-810. (c) Chang, A. Y.;Wyse, B. M.; Gilchrist, B. J.; Peterson, T.; Diani, A. R. Ciglitazone,a New HypoglycemicAgent. I. Studies in ob/ob and db/db Mice, Diabetic Chinese Hamsters, and Normal and Streptozotocin-Diabetic Rata. Diabetes 1983,32, 830-838.

0022-262319211835-2617$03.00/0 Q 1992 American Chemical society